Closed AlexandreBonneau closed 8 years ago
Well, after a few days investigating, it turns out it was a faulty thermal paste that did not conduct the heat from the CPU enough. It happened after having removed the sink to clean it...
In the first few days, sensors
nor hddtemp
did not show any alarming temperatures when I queried them, but today sensors
showed the CPU temperature skyrocket to 95 degrees Celsius a few seconds after starting to copy lots of data to the zfs array.
Since then, I changed the thermal paste and was able to copy 1.6Tio of data without problem.
my bad ;x
@AlexandreBonneau thanks for following up. Since it sounds like you got to the root cause I'm going to close this.
I experience a lot of system-wide crash under Debian 8.5 (ie. not responding to ping, ssh nor even keyboard/mouse inputs). This happens when I copy (a lot of) data from my old hdds to my brand new raid10 zol array.
Note : just to be sure, I stopped copying data to that pool, and for hours until now, I have had no problem whatsoever, whereas when I use my zol partition, I usually get a crash after 20 minutes or so.
Is that a known problem? Is there a log somewhere I could check apart from syslog? Is there a workaround?
Here is an extract of /var/log/syslog. You can see the bits of info before a crash, and until the second crash :